Neuronal networks must maintain constant performance for many years despite the fact that the ion channels and receptors that control excitability and synaptic transmission turn over in the membrane in hours, days, or weeks. Some neurological disorders and mental illnesses can be viewed as failures of network homeostasis, and therefore it becomes crucial to understand how stable network function is achieved over the lifetime of an animal. This proposal consists of combined experimental and computational work to ask whether network homeostasis can arise from cell autonomous regulation, or to what extent correlated activity measures or global feedback rules are required to maintain stable network function. The pyloric network of the crustacean stomatogastric nervous system is an ideal test-bed for these studies because the pyloric rhythm provides an easily measurable set of motor output patterns, its fictive motor patterns recorded in vitro resemble closely the motor patterns seen in vivo, and because a great deal is known about the identity of the neurons and their synaptic interactions. The proposed work will test the hypothesis that there is more animal-to-animal variance in the properties of single neurons and their synaptic connections than is found in the network output, because multiple combinations of synaptic strengths and intrinsic properties can be used by different animals to produce similar motor patterns. The proposed work includes quantitative analyses of pyloric motor patterns in vivo and in vitro, quantitative measurements of the intrinsic firing properties of single neurons, and quantitative measurements of synaptic strengths. Additionally it includes construction and analysis of databases of model neurons and model pyloric networks and computational studies of the tuning rules required to achieve network homeostasis.